bors [Fri, 28 Oct 2022 23:27:33 +0000 (23:27 +0000)]
Auto merge of #102737 - RalfJung:poll_fn_pin, r=Mark-Simulacrum
poll_fn and Unpin: fix pinning
See [IRLO](https://internals.rust-lang.org/t/surprising-soundness-trouble-around-pollfn/17484) for details: currently `poll_fn` is very subtle to use, since it does not pin the closure, so creating a `Pin::get_unchcked(&mut capture)` inside the closure is unsound. This leads to actual miscompilations with `futures::join!`.
IMO the proper fix is to pin the closure when the future is pinned, which is achieved by changing the `Unpin` implementation. This is a breaking change though. 1.64.0 was *just* released, so maybe this is still okay?
The alternative would be to add some strong comments to the docs saying that closure captures are *not pinned* and doing `Pin::get_unchecked` on them is unsound.
bors [Fri, 28 Oct 2022 20:19:28 +0000 (20:19 +0000)]
Auto merge of #2630 - RalfJung:windows-parking, r=RalfJung
Implement thread parking for Windows
Cc https://github.com/rust-lang/miri/issues/2628
Based on code by `@DrMeepster.` However I adjusted `WakeByAddressSingle`: I don't think the futex value is compared *again* after the thread is woken up. I see nothing in the Windows docs indicating such a comparison, and the Linux futex does not behave like that either. So we only check the value before sleeping, same as on Linux.
bors [Fri, 28 Oct 2022 16:27:56 +0000 (16:27 +0000)]
Auto merge of #103071 - wesleywiser:fix_inlined_line_numbers, r=davidtwco
Fix line numbers for MIR inlined code
`should_collapse_debuginfo` detects if the specified span is part of a
macro expansion however it does this by checking if the span is anything
other than a normal (non-expanded) kind, then the span sequence is
walked backwards to the root span.
This doesn't work when the MIR inliner inlines code as it creates spans
with expansion information set to `ExprKind::Inlined` and results in the
line number being attributed to the inline callsite rather than the
normal line number of the inlined code.
bors [Fri, 28 Oct 2022 09:41:40 +0000 (09:41 +0000)]
Auto merge of #103671 - matthiaskrgr:rollup-iuugpep, r=matthiaskrgr
Rollup of 5 pull requests
Successful merges:
- #102642 (Add tests for static async functions in traits)
- #103283 (Add suggestions for unsafe impl error codes)
- #103523 (Fix unwanted merge of inline doc comments for impl blocks)
- #103550 (diagnostics: do not suggest static candidates as traits to import)
- #103641 (Don't carry MIR location in `ConstraintCategory::CallArgument`)
Matthias Krüger [Fri, 28 Oct 2022 05:06:48 +0000 (07:06 +0200)]
Rollup merge of #103645 - weihanglo:update-cargo, r=weihanglo
Update cargo
6 commits in 9210810d1fd7b51ae0439a0a363cc50e36963455..7e484fc1a766f56dbc95380f45719698e0c82749 2022-10-25 22:31:50 +0000 to 2022-10-27 15:20:57 +0000
- fix(publish): Block until it is in index (rust-lang/cargo#11062)
- Add Accept-Encoding request header to enable compression (rust-lang/cargo#11292)
- Update contrib docs for highfive transition (rust-lang/cargo#11294)
- Migrate from highfive to triagebot (rust-lang/cargo#11293)
- Fix dupe word typos (rust-lang/cargo#11287)
- Fix confusing error messages when using -Zsparse-registry (rust-lang/cargo#11283)
Matthias Krüger [Fri, 28 Oct 2022 05:06:47 +0000 (07:06 +0200)]
Rollup merge of #103643 - notriddle:notriddle/summary-focus-visible, r=GuillaumeGomez
rustdoc: stop hiding focus outlines on non-rustdoc-toggle details tags
We really shouldn't be overriding this kind of stuff unless the browser default is really broken (like outlining the thing that isn't clickable). This directly reverts b8f4e74cbc938d3448507d422c98061c2b71c922.
Matthias Krüger [Fri, 28 Oct 2022 05:06:46 +0000 (07:06 +0200)]
Rollup merge of #103609 - BoxyUwU:fix_impl_self_cycle, r=compiler-errors
Emit a nicer error on `impl Self {`
currently it emits a "cycle detected error" but this PR makes it emit a more user friendly error specifically saying that `Self` is disallowed in that position. this is a pretty hacky fix so i dont expect this to be merged (I basically only made this PR because i wanted to see if CI passes)
Matthias Krüger [Fri, 28 Oct 2022 05:06:43 +0000 (07:06 +0200)]
Rollup merge of #103641 - compiler-errors:issue-103624, r=cjgillot
Don't carry MIR location in `ConstraintCategory::CallArgument`
It turns out that `ConstraintCategory::CallArgument` cannot just carry a MIR location in it, since we may bubble them up to totally different MIR bodies.
So instead, revert the commit a6b5f95fb028f9feb4a2957c06b35035be2c6155, and instead just erase regions from the original `Option<Ty<'tcx>>` that it carried, so that it doesn't ICE with the changes in #103220.
Best reviewed in parts -- the first is just a revert, and the second is where the meaningful changes happen.
We need this merge mechanism for inlined items but it's completely unwanted for impl blocks (at least the doc comments are, not the other attributes) since we want to keep what `cfg()` is put on the `pub use` or other attributes.
Matthias Krüger [Fri, 28 Oct 2022 05:06:42 +0000 (07:06 +0200)]
Rollup merge of #103283 - nbarrios1337:unsafe-impl-suggestions, r=cjgillot
Add suggestions for unsafe impl error codes
Adds suggestions for users to add `unsafe` to trait impls that should be `unsafe`, and remove `unsafe` from trait impls that do not require `unsafe`
With the folllowing code:
```rust
struct Foo {}
struct Bar {}
trait Safe {}
unsafe trait Unsafe {}
impl Safe for Foo {} // ok
impl Unsafe for Foo {} // E0200
unsafe impl Safe for Bar {} // E0199
unsafe impl Unsafe for Bar {} // ok
// omitted empty main fn
```
The current rustc output is:
```
error[E0199]: implementing the trait `Safe` is not unsafe
--> e0200.rs:13:1
|
13 | unsafe impl Safe for Bar {} // E0199
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^
error[E0200]: the trait `Unsafe` requires an `unsafe impl` declaration
--> e0200.rs:11:1
|
11 | impl Unsafe for Foo {} // E0200
| ^^^^^^^^^^^^^^^^^^^^^^
error: aborting due to 2 previous errors
Some errors have detailed explanations: E0199, E0200.
For more information about an error, try `rustc --explain E0199`.
```
With this PR, the future rustc output would be:
```
error[E0199]: implementing the trait `Safe` is not unsafe
--> ../../temp/e0200.rs:13:1
|
13 | unsafe impl Safe for Bar {} // E0199
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
help: remove `unsafe` from this trait implementation
|
13 - unsafe impl Safe for Bar {} // E0199
13 + impl Safe for Bar {} // E0199
|
error[E0200]: the trait `Unsafe` requires an `unsafe impl` declaration
--> ../../temp/e0200.rs:11:1
|
11 | impl Unsafe for Foo {} // E0200
| ^^^^^^^^^^^^^^^^^^^^^^
|
= note: the trait `Unsafe` enforces invariants that the compiler can't check. Review the trait documentation and make sure this implementation upholds those invariants before adding the `unsafe` keyword
help: add `unsafe` to this trait implementation
|
11 | unsafe impl Unsafe for Foo {} // E0200
| ++++++
error: aborting due to 2 previous errors
Some errors have detailed explanations: E0199, E0200.
For more information about an error, try `rustc --explain E0199`.
```
Matthias Krüger [Fri, 28 Oct 2022 05:06:41 +0000 (07:06 +0200)]
Rollup merge of #102642 - bryangarza:afit-tests, r=compiler-errors
Add tests for static async functions in traits
This patch adds test cases for AFIT, the majority of which are currently expected to run as `check-fail`.
---
Note: I grabbed the cases from https://hackmd.io/SwRcXCiWQV-WRJ4BYs53fA
Also, I'm not sure if the `async-associated-types2` and `async-associated-types2-desugared` are correct, I modified them a bit from the examples in the HackMD.
Michael Howell [Thu, 27 Oct 2022 20:36:27 +0000 (13:36 -0700)]
rustdoc: remove redundant `div.search-container`
This wrapper DIV was originally added in 89e1fb322321c05497caa01372ceb7d5b57fa680, when it allowed the search bar's
size to be calculated without using `calc()`. This `width` hack can be
removed using flexbox.
bors [Thu, 27 Oct 2022 21:36:07 +0000 (21:36 +0000)]
Auto merge of #103629 - matthiaskrgr:rollup-r94tqfa, r=matthiaskrgr
Rollup of 8 pull requests
Successful merges:
- #103110 (remove redundant Send impl for references)
- #103255 (Clean up hidden type registration)
- #103394 (Clarify documentation about the memory layout of `UnsafeCell`)
- #103408 (Clean return-position `impl Trait` in traits correctly in rustdoc)
- #103505 (rustdoc: parse self-closing tags and attributes in `invalid_html_tags`)
- #103524 (rustc_metadata: Add struct and variant constructors to module children at encoding time)
- #103544 (Add flag to forbid recovery in the parser)
- #103616 (rustdoc: remove CSS workaround for Firefox 29)
Weihang Lo [Thu, 27 Oct 2022 17:18:26 +0000 (01:18 +0800)]
Update cargo
6 commits in 9210810d1fd7b51ae0439a0a363cc50e36963455..7e484fc1a766f56dbc95380f45719698e0c82749
2022-10-25 22:31:50 +0000 to 2022-10-27 15:20:57 +0000
- fix(publish): Block until it is in index (rust-lang/cargo#11062)
- Add Accept-Encoding request header to enable compression (rust-lang/cargo#11292)
- Update contrib docs for highfive transition (rust-lang/cargo#11294)
- Migrate from highfive to triagebot (rust-lang/cargo#11293)
- Fix dupe word typos (rust-lang/cargo#11287)
- Fix confusing error messages when using -Zsparse-registry (rust-lang/cargo#11283)
Michael Howell [Thu, 27 Oct 2022 16:48:06 +0000 (09:48 -0700)]
rustdoc: stop hiding focus outlines on non-rustdoc-toggle details tags
We really shouldn't be overriding this kind of stuff unless the browser
default is really broken (like outlining the thing that isn't clickable).
This directly reverts b8f4e74cbc938d3448507d422c98061c2b71c922.
Matthias Krüger [Thu, 27 Oct 2022 13:03:58 +0000 (15:03 +0200)]
Rollup merge of #103616 - rust-lang:notriddle/moz-box-sizing, r=GuillaumeGomez
rustdoc: remove CSS workaround for Firefox 29
CSS variables, which rustdoc now relies on, are only supported in Firefox 31 and later: https://www.mozilla.org/en-US/firefox/31.0/releasenotes/
This means it’s fine to also rely on unprefixed box-sizing, which is supported in Firefox 29 and later: https://www.mozilla.org/en-US/firefox/29.0/releasenotes/
Matthias Krüger [Thu, 27 Oct 2022 13:03:58 +0000 (15:03 +0200)]
Rollup merge of #103544 - Nilstrieb:no-recovery-pls, r=compiler-errors
Add flag to forbid recovery in the parser
To start the effort of fixing #103534, this adds a new flag to the parser, which forbids the parser from doing recovery, which it shouldn't do in macros.
This doesn't add any new checks for recoveries yet and is just here to bikeshed the names for the functions here before doing more.
Matthias Krüger [Thu, 27 Oct 2022 13:03:56 +0000 (15:03 +0200)]
Rollup merge of #103394 - Pointerbender:unsafecell-docs, r=Amanieu
Clarify documentation about the memory layout of `UnsafeCell`
This PR addresses a [comment](https://github.com/rust-lang/rust/pull/101717#issuecomment-1279908390) by `@RalfJung` in PR #101717 to further clarify the documentation of `UnsafeCell<T>`. The previous PR was merged already before we had a chance to correct this, hence this second PR :)
To goal of this PR is:
1. Split the paragraph about the memory layout of `UnsafeCell<T>` and the usage of `UnsafeCell::(raw_)get()` into two paragraphs, so that it is easier to digest for the reader.
2. Slightly simplify the previously added examples in order to reduce redundancy between the new examples and the examples that already [existed](https://github.com/rust-lang/rust/blob/ddd119b2fed57eb6b19c44c18108de95c564a48d/library/core/src/cell.rs#L1858-L1908) before these 2 PRs (which remained untouched by both PRs).
Matthias Krüger [Thu, 27 Oct 2022 13:03:55 +0000 (15:03 +0200)]
Rollup merge of #103255 - oli-obk:opaque_wrong_eq_relation, r=compiler-errors
Clean up hidden type registration
work on https://github.com/rust-lang/rust/issues/101186
Actually passing down the relation and using it instead of `eq` for the hidden type comparison has *no* effect whatsoever and allows for no further improvements at the call sites. I decided the increased complexity was not worth it and thus did not include that change in this PR.
Matthias Krüger [Thu, 27 Oct 2022 13:03:55 +0000 (15:03 +0200)]
Rollup merge of #103110 - RalfJung:manual-send, r=thomcc
remove redundant Send impl for references
Also explain why the other instance is not redundant, move it next to the trait they are implementing, and out of the redundant module. This seems to go back all the way to https://github.com/rust-lang/rust/commit/35ca50bd5676db31a8074a216d1aadad7d434de8, not sure why the module was added.
The instance for `&mut` is the default instance we get anyway, and we don't have anything similar for `Sync`, so IMO we should be consistent and not have the redundant instance here, either.
bors [Thu, 27 Oct 2022 10:05:16 +0000 (10:05 +0000)]
Auto merge of #103623 - matthiaskrgr:rollup-318yc1t, r=matthiaskrgr
Rollup of 9 pull requests
Successful merges:
- #103035 (Even nicer errors from assert_unsafe_precondition)
- #103106 (Try to say that memory outside the AM is always exposed)
- #103475 (Make param index generation a bit more robust)
- #103525 (Move a wf-check into the site where the value is instantiated)
- #103564 (library: allow some unused things in Miri)
- #103586 (Process registered region obligation in `resolve_regions_with_wf_tys`)
- #103592 (rustdoc: remove redundant CSS selector `.notable-traits .notable`)
- #103593 (Remove an unused parser function (`Expr::returns`))
- #103611 (Add test for issue 103574)
Matthias Krüger [Thu, 27 Oct 2022 07:25:09 +0000 (09:25 +0200)]
Rollup merge of #103106 - saethlin:from_exposed_docs, r=thomcc
Try to say that memory outside the AM is always exposed
cc ``@Gankra`` ``@thomcc``
I want to confidently tell people that they can use `from_exposed_addr` to get a pointer for doing MMIO and/or other hardware interactions done with volatile reads/writes at particular addresses outside the Rust AM. Currently, the docs indicate that would be UB.
With this change, now the docs indicate that this is intended to be a valid use of `from_exposed_addr`.
Matthias Krüger [Thu, 27 Oct 2022 07:25:08 +0000 (09:25 +0200)]
Rollup merge of #103035 - saethlin:assert_unsafe_precondition3, r=thomcc
Even nicer errors from assert_unsafe_precondition
For example, now running `cargo test` with this patch I get things like:
```
$ cargo +stage1 test
Finished test [unoptimized + debuginfo] target(s) in 0.01s
Running unittests src/lib.rs (target/debug/deps/malloc_buf-9d105ddf86862995)
running 5 tests
thread 'tests::test_null_buf' panicked at 'unsafe precondition violated: is_aligned_and_not_null(data) &&
crate::mem::size_of::<T>().saturating_mul(len) <= isize::MAX as usize', /home/ben/rust/library/core/src/slice/raw.rs:93:9
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
thread panicked while panicking. aborting.
error: test failed, to rerun pass `--lib`
Caused by:
process didn't exit successfully: `/tmp/malloc_buf-1.0.0/target/debug/deps/malloc_buf-9d105ddf86862995` (signal: 6, SIGABRT: process abort signal)
```
This is still not perfect, but these are better for another PR:
* `stringify!` is trying to do clever pretty-printing on the `expr` inside `assert_unsafe_precondition` and can even add a newline.
* It would be nice to print a bit more information about where the problem is. Perhaps this is `cfg_attr(debug_assertions, track_caller)`, or perhaps it the function name added to `Location`.
cc ``@RalfJung`` this is what I was thinking of for https://github.com/rust-lang/rust/pull/102732#discussion_r989068907
bors [Thu, 27 Oct 2022 07:24:39 +0000 (07:24 +0000)]
Auto merge of #103601 - compiler-errors:no-opaque-probe-in-nll-relate, r=oli-obk
Remove `commit_if_ok` probe from NLL type relation
It was not really necessary to add the `commit_if_ok` in #100092 -- I added it to protect us against weird inference error messages due to recursive RPIT calls, but we are always on the error path when this happens anyways, and I can't come up with an example that makes this manifest.
Fixes #103599
r? `@oli-obk` since you reviewed #100092, feel free to re-roll.
:b: :loudspeaker: beta-nominating this since it's on beta (which forks in ~a week~ two days :fearful:) -- worst case we could revert the original PR on beta and land this on nightly, to give it some extra soak time...
Michael Howell [Thu, 27 Oct 2022 04:13:20 +0000 (21:13 -0700)]
rustdoc: remove CSS workaround for Firefox 29
CSS variables, which rustdoc now relies on, are only supported in Firefox 31 and later: https://www.mozilla.org/en-US/firefox/31.0/releasenotes/
This means it’s fine to also rely on unprefixed box-sizing, which is supported in Firefox 29 and later: https://www.mozilla.org/en-US/firefox/29.0/releasenotes/
projects / rust.git / log